Many systems that use entity beans to hold instances of data for an application will generate a separate cache for each entity bean. It is then necessary to configure each of these caches. Utilizing separate caches can lead to problems with memory fragmentation. Further, users can set limits on the size of these individual caches such that the system may be unable to use all available memory. For example, one of the entity beans might be very busy and require a lot of memory, while other beans sit idle. If the caches are configured to accommodate the busiest bean, the unused beans will have large cache allocations that will not be utilized, thereby wasting memory.